In 2010, it was estimated that type 2 diabetes mellitus (T2DM) affects nearly 26 million individuals in the U.S. or 8.3% of the population. The US Centers for Disease Control and Prevention released alarming figures in January 2011 showing that the number of American adults with prediabetes, a condition associated with a high risk of developing frank T2DM, had increased from 57 million in 2008 to 79 million in 2010. The prevalence of obstructive sleep apnea (OSA) in the general population has also increased rapidly and dramatically, in parallel with the epidemic of obesity. In fact, the prevalence of OSA in patients with T2DM ranges between 58 to 86% making it the most common comorbidity in T2DM. It has been estimated that the prevalence of OSA is also very high in prediabetes. In addition to the obesity epidemic, self-reported sleep duration in Americans has decreased by 1.5-2 h over the last few 4 decades. Over the past decade, both laboratory and epidemiologic studies have identified insufficient sleep, sleep fragmentation, poor sleep quality and OSA, as putative novel risk factors for T2DM. Numerous studies have established a robust association between the presence and severity of OSA with both insulin resistance and glucose intolerance, independent of adiposity. However, to date, not a single clinical trial of lifestyle and/or pharmacological interventions for prediabetes or T2DM has taken into account the role of sleep disturbances despite their very high prevalence in prediabetes and T2DM and their intrinsic adverse impact on insulin release and action. It is not known whether individuals suffering from sleep disturbances are more resistant than those without sleep disruption to pharmacological treatment for prediabetes or T2DM. The parent RISE Study was designed to test the impact of intensive short-term pharmacological interventions to preserve or enhance beta cell function to prevent or delay the development of frank T2DM. The overall goals of the proposed ancillary study are: 1) To determine if quantitative measures of sleep quantity and quality are predictors of baseline measures of -cell function (insulin secretion) and insulin sensitivity, and 2) To assess whether the presence and severity of sleep disturbances impact the -cell response to pharmacological interventions designed to preserve and/or restore insulin secretion. This proposal will capitalize on the RISE Study's uniquely detailed quantification of -cell function and insulin sensitivity in more than 250 participants both at baseline and after randomization to an intervention. Prior to randomization, participants will undergo one week of home wrist actigraphy and one night of in-laboratory polysomnography (PSG). We hypothesize that in subjects with prediabetes or early T2DM, habitual short sleep duration, poor habitual sleep quality, presence and severity of OSA and low levels of SWA are each associated with reduced -cell function and decreased insulin sensitivity. We also hypothesize that each of these sleep disturbances will decrease the effectiveness of pharmacological interventions intended to preserve and restore -cell function.